What type of study is this?

September 10, 2025Structural Health MonitoringOpen Access

An interpretable transfer learning method for bearing diagnosis across different systems, faults, and signal types

View Full Paper

Ask AI

Bookmark

Authors

This approach improves diagnostic accuracy for bearing faults in varying signal types, indicating enhanced adaptability.

The method achieves effective transfer learning, yielding high accuracy across different machines and fault types.
Case study results show that the transfer learning technique enhances model performance with limited labeled data.
The SHAP analysis shows that key time-frequency features related to bearing faults can be effectively identified.
Integrating feature reshaping improves the model's signal-to-noise ratio, focusing on relevant fault features.

Rong et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark